What Is the Theory of Constraints?
The Theory of Constraints, developed by Eli Goldratt and popularized in his book The Goal, is a management philosophy built on one insight: every system has exactly one constraint that limits its total output. Improving anything other than the constraint does not improve the system — it just creates more WIP, more confusion, and the illusion of progress.
In a manufacturing plant, the constraint is the bottleneck — the process step with the least capacity. If your bottleneck can produce 80 units per hour, your system produces 80 units per hour, regardless of whether every other step can do 200.
The Most Dangerous Improvement
Improving a non-bottleneck is worse than doing nothing. If you speed up a step upstream of the bottleneck, you just pile up more WIP in front of it. If you speed up a step downstream, it starves faster. The only improvement that increases system output is improving the constraint. Everything else is a costly distraction.
Both rates are the page's own: "if your bottleneck can produce 80 units per hour, your system produces 80 units per hour, regardless of whether every other step can do 200". The gap is 200 − 80 = 120 units an hour, and it is the same 120 in all three panels. The only thing added here is a window to accumulate it over — eight hours, because the page gives rates per hour and never names a shift length — so 120 × 8 = 960 units. Shipped stays 80 × 8 = 640 units in the first two panels, because the constraint has not moved. Panel 1: the upstream step releases at its own 200/hr while the constraint draws 80/hr, so 120/hr piles up — 960 units of WIP by hour 8, half again more than everything shipped. Panel 2: the downstream step is fed 80/hr and can do 200/hr, so it works 80/200 = 40% of the time and stands idle 4.8 of the 8 hours; the capacity it offers with nothing to work on is the same 960 units. Panel 3: the constraint now runs at the 200/hr every other step can already do, so the system ships 200 × 8 = 1,600, and 1,600 − 640 = 960. T and I are the page's own throughput accounting — throughput is the rate at which the system generates money, maximise it; inventory includes WIP, minimise it.
The 5 Focusing Steps
Goldratt's framework for systematic constraint management:
Drum-Buffer-Rope (DBR)
DBR is the TOC scheduling method that synchronizes the entire plant to the constraint:
| Element | What It Is | Purpose |
|---|---|---|
| Drum | The constraint's production schedule | Sets the pace for the entire system. Everything marches to this beat. |
| Buffer | A time buffer of WIP in front of the constraint | Protects the constraint from upstream disruptions. If an upstream machine breaks, the buffer keeps the bottleneck fed. |
| Rope | A signal to the first operation to release material | Controls the rate of material entry into the system. Material is released at the rate the constraint can process it — no faster. This prevents WIP explosion. |
(Constraint)
Throughput Accounting
TOC uses a different financial lens than traditional cost accounting:
| Metric | Definition | Key Insight |
|---|---|---|
| Throughput (T) | Revenue – truly variable costs (materials) | The rate at which the system generates money. Maximize this. |
| Inventory (I) | All money invested in things intended for sale | Includes raw material, WIP, finished goods, equipment. Minimize this. |
| Operating Expense (OE) | All money spent to turn inventory into throughput | Labor, rent, utilities, overhead. Control this. |
The TOC Priority
Traditional accounting focuses on cutting OE (layoffs, cost reduction). TOC says: increase T first, decrease I second, then control OE. A $1 increase in throughput by exploiting the constraint is worth far more than a $1 cut in operating expense, because throughput is theoretically unlimited while cost cutting has a floor of zero.
Finding Your Constraint
| Clue | What It Tells You |
|---|---|
| WIP piling up in front of a process | That process is likely the constraint — it cannot keep up |
| Downstream processes starving for work | Something upstream is not producing fast enough |
| One machine always running, others idle | The always-running machine is the constraint |
| Customer lead times driven by one process | That process is the constraint for delivery |
| Overtime only needed at one process step | Capacity-constrained at that step |
✅ TOC Thinking
- Identify the ONE constraint and focus all energy there
- Exploit before spending: maximize output with existing resources
- Subordinate: non-bottlenecks run at bottleneck pace
- Measure throughput, not local efficiencies
- When constraint moves, follow it
❌ Local Optimization Trap
- Improve every machine's efficiency regardless of bottleneck
- Run all machines at maximum speed (creates WIP mountains)
- Measure each department on utilization (drives overproduction)
- Invest capital at non-constraints
- Celebrate efficiency gains that do not increase shipments
🎯 Key Takeaway
Your plant's output is determined by one constraint. Find it, exploit it, subordinate everything else to it, and only then invest to elevate it. Improving non-constraints feels productive but does not increase a single shipment. The elegance of TOC is its focus: stop trying to improve everything and pour all your energy into the one place that matters. Combine with OEE tracking on the bottleneck and SMED to exploit it fully.
Find the Bottleneck
Adjust cycle times for each station and watch the bottleneck shift. System throughput always equals the bottleneck's throughput — try improving it and see what happens.
Stop reading, start modeling
Model your process flow, run simulations, optimize staffing with TOC math, and test your knowledge with 107 interactive checks — all in one platform.